A Qualitative Transcriptional Signature for Predicting Microsatellite Instability Status of Right-sided Colon Cancer

doi:10.21203/rs.2.10681/v1

Download PDF

Research article

A Qualitative Transcriptional Signature for Predicting Microsatellite Instability Status of Right-sided Colon Cancer

https://doi.org/10.21203/rs.2.10681/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 23 Oct, 2019

Read the published version in BMC Genomics →

You are reading this older preprint version

Read the latest preprint version →

Background: Microsatellite instability (MSI) accounts for about 15% of colorectal cancer and is associated with prognosis. Today, MSI is usually detected by polymerase chain reaction amplification of specific microsatellite markers. However, the instability is identified by comparing the length of microsatellite repeats in tumor and normal samples. In this work, we developed a qualitative transcriptional signature to individually predict MSI status for right-sided colon cancer (RCC) based on tumor samples. Results: Using RCC samples, based on the relative expression orderings (REOs) of gene pairs, we extracted a signature consisting of 10 gene pairs (10-GPS) to predict MSI status for RCC through a feature selection process. A sample is predicted as MSI when the gene expression orderings of at least 7 gene pairs vote for MSI; otherwise the microsatellite stability (MSS). The classification performance reached the largest F-score in the training dataset. This signature was verified in four independent datasets of RCCs with the F-scores of 1, 0.9630, 0.9412 and 0.8798, respectively. Additionally, the hierarchical clustering analyses and molecular features also supported the correctness of the reclassifications of the MSI status by 10-GPS. Conclusions: The qualitative transcriptional signature can be used to classify MSI status of RCC samples at the individualized level.

Cancer Biology

Epigenetics & Genomics

right-sided colon cancer

microsatellite instability status

gene expression profiles

relative gene expression orderings

qualitative transcriptional signature

Microsatellite instability (MSI), the insertion or deletion mutations in microsatellites [ 1], is a molecular hallmark of a deficient mismatch repair (dMMR) system and accounts for about 15% of colorectal cancer (CRC) [ 2]. Results from some studies seem that the MSI feature is associated with good prognosis, and the stage II and III CRC patients with MSI cannot benefit from 5-fluorouracil (5-Fu)-based adjutant chemotherapy (ACT) [ 3-5], which is regarded as the standard treatment for stage II and III CRC patients after surgery. So, a precise classification is needed to aid appropriate decisions on 5-Fu-based ACT treatment of patients.

Today, the most common method to test MSI status is polymerase chain reaction (PCR) amplification analysis of specific microsatellite repeats, which is considered as the 'golden standard' method [ 1, 6]. However, the PCR technology exists high measurement variations between different laboratories [ 7, 8], which are mainly due to the effects of the tumor cell percentage and DNA degradation during sample storage and preparation [ 8-11]. Additionally, the presence of instability is defined by comparing the length of microsatellite repeats in the tumor sample and the normal sample [ 1, 6, 12-14]. MSI can also be detected by immunohistochemistry (IHC). But IHC method only provides a semi-quantitative evaluation of the expression levels of the four MMR proteins (MLH1, MSH2, MSH6 and PMS2) [ 2], and also exists high measurement variations between different laboratories [ 15-17], which are primarily due to the effects of sample preprocessing, such as fixation of tissues, detection reagents and selection of antibody [ 16, 17]. Moreover, since the results of IHC can be greatly affected by the interpretation of the specificity of staining, when the levels of the MMR proteins analyzed are low, good performance of IHC requires highly skilled personnel and pathologist’s experience in interpretation [ 14, 15]. Therefore, the traditional PCR and IHC methods both have some limitations in determining the MSI status of CRC.

Recently, many alternative methods based on tumor genomic data via next-generation sequencing (NGS) panels have been developed to determine MSI status of patients[ 1, 18, 19]. For instance, Vanderwalde et al. used a NGS panel comprised of 592 genes to determine MSI status [ 1]. However, NGS currently has been limited to some highly specialized laboratories [ 20]. Besides, there has no consensus about NGS gene panels to determine MSI status and each laboratory determine the appropriate mutation load threshold based on its NGS gene panel and technique [ 1, 18-20]. What’s more, the use of the NGS-based methods that require DNA extraction often leads to false-negative or uncertain results in challenging tumor samples due to the tumor DNA dilution and the percentage of tumor cells within a sample [ 20, 21]. Another method based on gene expression measurement also has been developed to identify MSI status of CRCs [ 6], which, however, is sensitive to the systematic inter-laboratory biases especially batch effects of microarray and RNA-sequencing experiments [ 22]. In general, quantitative transcriptional signatures based on absolute expression values is sensitive to the batch effects and thus lack robustness for clinical applications [ 23]. In contrast, the type of qualitative transcriptional signatures based on the within-sample relative expression orderings (REOs) of genes have strong robustness against the experimental batch effects and can be applied to individual samples directly [ 24]. Besides, we have demonstrated that they are rather robust against partial RNA degradation during sample preparation [ 25], the proportions of tumor epithelial cell variations in tumor tissues sampled from different tumor locations [ 26], and amplification bias for minimum samples [ 27]. Therefore, it is worthwhile to apply the within-sample REOs to find robust qualitative transcriptional signatures.

Colorectal cancers deriving from proximal or distal of splenic flexure are classified as right-sided or left-sided colon cancer (RCC or LCC), respectively [ 28]. Consistent with the differences in anatomy location, RCC and LCC have unique gene expression characteristics, different molecular pathways of carcinogenesis [ 28, 29] and different clinical features [ 30]. Therefore, it would be necessary to develop signatures to predict MSI status for RCC and LCC, respectively. Because of the high incidence of MSI in the RCC, we developed a REOs-based qualitative signature for predicting MSI status of RCC patients in this work, which was validated in independent datasets.

Identification and validation of the signature for MSI status of RCC

The GSE39582 with the largest samples of RCC, including 57 MSI and 154 MSS was used as the training data for extracting a REOs-based signature. Firstly, we identified 4769 MSI-related DE genes (Student's t-test, FDR < 0.01) between the 57 MSI RCC samples and the 154 MSS RCC samples. From all the gene pairs formed by these DE genes, we identified 1,654,739 gene pairs, whose specific REO pattern occurred more frequently in the MSI samples than in the MSS samples (Fisher’s exact test, FDR < 0.01). The larger FD of a REO pattern, the stronger classified ability of the REO pattern can classify the status of MSI. We further narrowed down the number of gene pairs to 1898 through a redundancy removal process by keeping only one with the largest FD value of those gene pairs sharing a common gene (see Methods, Figure 1a). From these gene pairs, we extracted 10 gene pairs with the FD at least 0.8. These 10 gene pairs (Table 1), denoted as 10-GPS, were used as the signature for predicting MSI status of RCC. A RCC sample was predicted as MSI if the REOs of at least seven gene pairs in the 10-GPS vote for MSI; otherwise the MSS. According to the classification rule, the F-score of the signature in the training data was 0.9727, with a sensitivity of 0.9649 and a specificity of 0.9805. The area under the curve (AUC) of the receiver operating characteristic (ROC) curve was 0.9838 (Figure 2a).

Table 1. The Composition of 10-GPS

We tested the 10-GPS in four independent cohorts of RCC samples (Figure 1b), the F-scores of the classification by 10-GPS were 1, 0.9630, 0.9412 and 0.8798, respectively, as shown in the Table 2, and the AUCs were 1, 0.9923, 1 and 0.9244, respectively (Figure 2b, c, d and e).

Table 2. The performances of the 10-GPS in RCCs of the independent datasets

Transcriptome assessment of the signature-disconfirmed RCC samples

In the training data, there were a total of five signature-disconfirmed samples. We compared the gene expression patterns of the five signature-disconfirmed samples with the 206 signature-confirmed samples through clustering analysis. Firstly, we identified 5664 DE genes (Student's t-test, FDR < 0.01) between the 55 signature-confirmed MSI and the 151 signature-confirmed MSS samples. Secondly, using the expression levels of the top 100 significant DE genes, the samples were divided into two subgroups using the complete linkage hierarchical clustering based on the Euclidean distance (Figure 3a). The results showed that all of the two MSI samples reclassified as MSS by the 10-GPS were clustered with the signature-confirmed MSS samples, and all of the three MSS samples reclassified as MSI were clustered with the signature-confirmed MSI samples.

Similarly in the two of the four validation datasets of RCC (GSE18088 and GSE75317), all of these two MSI samples reclassified as MSS by our signature were clustered with the corresponding signature-confirmed MSS samples (Figure 3b and c), respectively. These results provided transcriptional evidence of the correctness of the prediction of 10-GPS.

Genome assessment of the signature-disconfirmed RCC samples

It is known that BRAF^V600E mutations and CpG island methylator phenotype (CIMP)-positive frequently occur in MSI CRCs, whereas KRAS mutations (in codons 12 or 13) frequently occur in MSS CRCs [ 2, 31]. In the training data, for the three MSS samples which were reclassified as MSI by 10-GPS, two patients were KRAS wild-type, BRAF mutant and CIMP-positive and one patient was KRAS wild-type. For the two MSI samples which were reclassified as MSS, one was BRAF wild-type and CIMP-negative, as shown in Table 3.

Table 3. The molecular characteristics of the five signature-disconfirmed RCC samples in the training dataset

In the TCGA validation dataset of RCC, there were seven signature-disconfirmed samples. Because mutation of MMR genes can result in MSI [ 32], we observed the mutation status of the MMR genes in the signature-disconfirmed samples. There were only five samples with mutation data. And two of the four MSS samples which were reclassified as MSI by 10-GPS were MSH6 mutant (Supplementary Table 1, Additional File 1). These results supported that MSI status of these samples reclassified by 10-GPS might be reliable.

Prognosis assessment of the signature-disconfirmed RCC samples

Then, we also evaluated the reliability of the reclassifications by 10-GPS through survival analyses based on the knowledge that stage III MSI CRCs treated with surgery only have better prognoses than MSS CRCs [ 2] and that stage III MSS CRC patients treated with 5-Fu-based ACT after surgery have improved outcomes than patients treated with surgery only [ 3-5]. And the survival benefit of ACT was only observed in stage III patients [ 3, 4]. In the 32 stage III RCC samples of the training data for patients treated with surgery only, one of the 19 MSS sample was reclassified as MSI and one of 13 MSI sample was reclassified as MSS by the 10-GPS. In the 54 (46 MSS and 8 MSI) stage III RCC samples for patients receiving 5-Fu-based ACT, the original MSI status of all samples were confirmed by the 10-GPS. By comparison, the MSS patient reclassified as MSI had longer RFS (130 months) than the MSI patient reclassified as MSS (31 months) by the signature. The survival difference between patients with predicted MSI status by 10-GPS were more significant than the difference between patients with the original MSI status due to the two reclassified samples (Supplementary Figure, Additional File 2).

In this study, we developed a qualitative transcriptional signature consisting of 10 gene pairs to robustly predict MSI status of RCC at individualized level, which was validated in four independent datasets. Notably, the hierarchical clustering analyses and molecular characteristics supported the correctness of the reclassifications of the MSI status by our signature for some samples whose MSI statuses were determined by the PCR testing. Thus, it is possible for our signature to identify the MSI status which could not be determined by traditional standard methods.

We applied the 10-GPS to LCC samples and CRC samples without clear location information (Supplementary Table 2, Additional File 1). The results showed that the performance was reduced when applying the 10-GPS to predict MSI status of LCC. Therefore, we also tried to develop a signature to identify MSI status of the LCC patients in the same way as in RCC. Eventually, these six gene pairs (Supplementary Table 3, Additional File 1), denoted as 6-GPS, were used as the signature for predicting MSI status of LCC. A LCC sample was predicted as MSI if the REOs of at least four gene pairs in the 6-GPS vote for MSI; otherwise the MSS. According to the classification rule, the F-score of the signature in the LCC training data was 0.9983, with a sensitivity of 1 and a specificity of 0.9966. And, the 6-GPS was well validated in four independent cohorts of LCC samples (Supplementary Table 4, Additional File 1).

In the process of screening MSI-related gene pairs of RCC, we extracted gene pairs by adjusting different FD thresholds. Then, it was found that when FD > 0.9, there were no gene pairs remaining. When FD > 0.7, there were 65 gene pairs were extracted. The classification performance reached the largest F-score (0.9630) of sensitivity (0.9649) and specificity (0.9610) in the training data according to the following decision rule: a sample is predicted as MSI when the REOs of at least 39 gene pairs vote for MSI; otherwise the MSS. Compared with the result of FD > 0.8, the performance of classification was slightly worse and it had more gene pairs. So we chose 0.8 as the threshold for FD. Similarly, based on the same considerations, we chose 0.9 as the threshold for FD in the process of screening MSI-related gene pairs of LCC.

Given cost considerations and the often-limited quantity of tumor material available for testing in many cancer patients, NGS-based tumor profiling represents the practical method to assess for the presence of the numerous standard and investigational biomarkers of potential interest in an individual patient [ 33], which provides the basis for the concept of “a sequence for all”. For example, we have developed a series of qualitative transcriptional signatures for the prognosis prognostic and prediction predictive biomarkers of stage II-III CRC, including predicting histological grade, CIMP status, micro-metastasis after surgery [ 34] and subsequent 5-FU-based ACT benefit of high relapse risk patients [ 35]. All these signatures can be assessed in a single NGS assay, facilitating the optimum treatment of stage II-III CRC patients. Herein, we developed qualitative signatures for predicting MSI status of RCC and LCC, as a part of “a sequence for all” for CRC.

In summary, we developed two qualitative transcriptional signatures for predicting MSI status of RCC and LCC, respectively. Particularly, the two signatures are simple and robust against the experimental batch effects, the proportions of the tumor epithelial cell variations in tumor tissues sampled from different tumor locations, partial RNA degradation during sample preparation, and amplification bias for minimum samples [ 25-27] and do not need matched normal tissues. What’s more, the qualitative signature can be applied at the individual level directly, which avoids the impractical requirement of pre-collecting a group of samples for data normalization [ 36].

Conclusions

Currently, common methods for detecting MSI status of CRC such as PCR and IHC-based methods, exist high measurement variations between different laboratories, which have limited clinical utility. Herein, we developed a robust qualitative transcriptional signature to classify MSI status of RCC at the individualized level, as a part of “a sequence for all” for CRC. The simplicity and robustness of the signature would make it convenient in clinical settings and worthy to further validate in a prospective clinical trial.

Data sources and data preprocessing

The gene expression datasets used in this study were downloaded from the Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/) database and The Cancer Genome Atlas data portal (TCGA, http://cancergenome.nih.gov/) (Table 4). As MSI-L patients are usually treated in a way similar to MSS patients in clinical practice, it is reasonable to group MSI-L with MSS [ 1]. In this study, we grouped MSI-L with MSS. The training data for extracting a REOs-based signature of RCC was GSE39582, including 57 MSI and 154 MSS of 211 RCCs. The GSE39582 dataset recording survival information of patients were used as the test for survival analyses.

The Robust Multi-array Average algorithm [ 37] was used for preprocessing the raw data measured by the Affymetrix platform. Using the corresponding platform files, probes were mapped to genes. For each sample, the expression measurements of several probes mapping to a gene were averaged to obtain a single measurement. Probes were discarded if they did not match any gene or matched multiple genes. The RNA-seq expression data were downloaded from the Broad Firehose webpage (http://gdac.broadinstitute.org/). For RNA-seq data derived from Illumina HiSeq 2000 RNA Sequencing Version 2, we directly downloaded the RSEM-normalized format and log2-transformed. We also downloaded the somatic mutation data of CRC from the Broad Firehose webpage.

Table 4 is about here.

Currently, most of the CRC data we collected were microarray datasets. With the development of the NGS technology, the RNA-seq data are increasing. In order to apply our signature to RNA-seq data, we used the expression data of overlapping Gene IDs of Affymetrix and Illumina platforms.

Signature development for predicting MSI status of RCC

We identified MSI-related gene pairs from the DE genes between MSI and MSS samples. For a gene pair, i and j, with expression values of E_iand E_j, whether the frequency of a specific REO pattern (E_i> E_j or E_i< E_j) was significantly higher in the MSI samples than the frequency in the MSS samples was evaluated by Fisher’s exact test [ 38]. The MSI-related gene pairs detected with FDR < 0.01. Then, the frequency difference (FD) was calculated for each MSI-related gene pair between the two groups (Formula 1). For a gene which appeared in multiple gene pairs, we kept only the gene pair with the largest FD value and discarded others.

p_ij(c) = P(E_I > E_j |c), c = 1,2, the probabilities of observing E_i> E_j in each class

FD_ij = p_ij(1) - p_ij(2), the FD value of gene pair (i, j) [Formula 1]

After that, the gene pairs with the FD value at least 0.8 were identified as the signature to predict MSI status of RCC. A sample was labeled as MSI if the REOs of at least k gene pairs in the signature are consistent with the specific patterns (E_i > E_j) of the training samples, and vice versa. For each k ranging from one to the number of gene pairs in the signature, we could compute a F-score. Finally, we selected the k which could reach the largest F-score.

F - score = 2 x sensitivity x specificity ÷ (sensitivity + specificity) [Formula 2]

Sample clustering

Student’s t-test was performed to identify DE genes between the MSI and MSS patients confirmed with the original MSI status by the signature. Complete linkage hierarchical clustering was performed to stratify CRC samples into subgroups. The similarity between samples was evaluated by the Euclidean distance based on the expression measurements of DE genes.

Statistical analyses

For microarray data, we selected differentially expressed (DE) genes between two classes of samples using Student’s t-test. The MSI rate in different groups was evaluated by Fisher’s exact test. The Benjamini-Hochberg procedure for multiple testing was used to adjust p values in order to control the false discovery rate (FDR) [ 39]. Log-rank test was used to assess the difference between the Kaplan-Meier estimates of the RFS in the two different groups [ 40]. Hazard ratios (HRs) and 95% confidence intervals (CIs) were generated using the univariate Cox proportional hazards model. All statistical analyses were performed using the R 3.5.1 (http://www.r-project.org/).

MSI: Microsatellite instability; RCC: right-sided colon cancer; REOs: relative expression orderings; FD: frequency difference; GPS: gene pairs; MSS: microsatellite stability; CRC: colorectal cancer; 5-Fu: 5-fluorouracil; ACT: adjutant chemotherapy; PCR: polymerase chain reaction; IHC: immunohistochemical; NGS: next-generation sequencing; AUC: area under the curve; ROC: receiver operating characteristic; CIMP: CpG island methylator phenotype; LCC: left-sided colon cancer; dMMR: deficient mismatch repair; DE: differentially expressed; FDR: false discovery rate.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and materials

All data analyzed in this study were downloaded from the public database: GEO and TCGA.

Competing interests

The authors declare that they have no competing interests.

Funding

This work was supported by the National Natural Science Foundation of China [grant numbers: 61601151, 61701143, 61673143, 81872396 and 81572935], the Natural Science Foundation of Heilongjiang Province [grant number: C2016037] and the Joint Scientific and Technology Innovation Fund of Fujian Province [grant number: 2016Y9044].

Authors' contributions

WZ and ZG conceived the idea, YF conceived and designed the experiments, wrote the manuscript, LQ designed the experiments, WG and LJ analyzed the data, KS and TY performed the experiments, SZ and YG helped in writing the manuscript. All authors approved the final version.

Acknowledgements

Not applicable.

1. Vanderwalde A, Spetzler D, Xiao N, Gatalica Z, Marshall J: Microsatellite instability status determined by next-generation sequencing and compared with PD-L1 and tumor mutational burden in 11,348 patients. Cancer medicine 2018, 7(3):746-756.

2. Gelsomino F, Barbolini M, Spallanzani A, Pugliese G, Cascinu S: The evolving role of microsatellite instability in colorectal cancer: A review. Cancer treatment reviews 2016, 51:19-26.

3. Cao B, Luo L, Feng L, Ma S, Chen T, Ren Y, Zha X, Cheng S, Zhang K, Chen C: A network-based predictive gene-expression signature for adjuvant chemotherapy benefit in stage II colorectal cancer. BMC cancer 2017, 17(1):844.

4. Sargent DJ, Marsoni S, Monges G, Thibodeau SN, Labianca R, Hamilton SR, French AJ, Kabat B, Foster NR, Torri V et al: Defective mismatch repair as a predictive marker for lack of efficacy of fluorouracil-based adjuvant therapy in colon cancer. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 2010, 28(20):3219-3226.

5. Ribic CM, Sargent DJ, Moore MJ, Thibodeau SN, French AJ, Goldberg RM, Hamilton SR, Laurent-Puig P, Gryfe R, Shepherd LE et al: Tumor microsatellite-instability status as a predictor of benefit from fluorouracil-based adjuvant chemotherapy for colon cancer. The New England journal of medicine 2003, 349(3):247-257.

6. Tian S, Roepman P, Popovici V, Michaut M, Majewski I, Salazar R, Santos C, Rosenberg R, Nitsche U, Mesker WE et al: A robust genomic signature for the detection of colorectal cancer patients with microsatellite instability phenotype and high mutation frequency. The Journal of pathology 2012, 228(4):586-595.

7. Hempelmann JA, Lockwood CM, Konnick EQ, Schweizer MT, Antonarakis ES, Lotan TL, Montgomery B, Nelson PS, Klemfuss N, Salipante SJ et al: Microsatellite instability in prostate cancer by PCR or next-generation sequencing. Journal for immunotherapy of cancer 2018, 6(1):29.

8. Boyle TA, Bridge JA, Sabatini LM, Nowak JA, Vasalos P, Jennings LJ, Halling KC, College of American Pathologists Molecular Oncology C: Summary of microsatellite instability test results from laboratories participating in proficiency surveys: proficiency survey results from 2005 to 2012. Archives of pathology & laboratory medicine 2014, 138(3):363-370.

9. Waalkes A, Smith N, Penewit K, Hempelmann J, Konnick EQ, Hause RJ, Pritchard CC, Salipante SJ: Accurate Pan-Cancer Molecular Diagnosis of Microsatellite Instability by Single-Molecule Molecular Inversion Probe Capture and High-Throughput Sequencing. Clinical chemistry 2018, 64(6):950-958.

10. Chen G, Yang Z, Eshleman JR, Netto GJ, Lin MT: Molecular Diagnostics for Precision Medicine in Colorectal Cancer: Current Status and Future Perspective. BioMed research international 2016, 2016:9850690.

11. Trusky CL, Sepulveda AR, Hunt JL: Assessment of microsatellite instability in very small microdissected samples and in tumor samples that are contaminated with normal DNA. Diagnostic molecular pathology : the American journal of surgical pathology, part B 2006, 15(2):63-69.

12. Ryan E, Sheahan K, Creavin B, Mohan HM, Winter DC: The current value of determining the mismatch repair status of colorectal cancer: A rationale for routine testing. Critical reviews in oncology/hematology 2017, 116:38-57.

13. Zeinalian M, Hashemzadeh-Chaleshtori M, Salehi R, Emami MH: Clinical Aspects of Microsatellite Instability Testing in Colorectal Cancer. Advanced biomedical research 2018, 7:28.

14. Pena-Diaz J, Rasmussen LJ: Approaches to diagnose DNA mismatch repair gene defects in cancer. DNA repair 2016, 38:147-154.

15. Shia J, Holck S, Depetris G, Greenson JK, Klimstra DS: Lynch syndrome-associated neoplasms: a discussion on histopathology and immunohistochemistry. Familial cancer 2013, 12(2):241-260.

16. Kirkegaard T, Edwards J, Tovey S, McGlynn LM, Krishna SN, Mukherjee R, Tam L, Munro AF, Dunne B, Bartlett JM: Observer variation in immunohistochemical analysis of protein expression, time for a change? Histopathology 2006, 48(7):787-794.

17. Fitzgibbons PL, Murphy DA, Hammond ME, Allred DC, Valenstein PN: Recommendations for validating estrogen and progesterone receptor immunohistochemistry assays. Archives of pathology & laboratory medicine 2010, 134(6):930-935.

18. Stadler ZK, Battaglin F, Middha S, Hechtman JF, Tran C, Cercek A, Yaeger R, Segal NH, Varghese AM, Reidy-Lagunes DL et al: Reliable Detection of Mismatch Repair Deficiency in Colorectal Cancers Using Mutational Load in Next-Generation Sequencing Panels. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 2016, 34(18):2141-2147.

19. Salem ME, Puccini A, Grothey A, Raghavan D, Goldberg RM, Xiu J, Korn WM, Weinberg BA, Hwang JJ, Shields AF et al: Landscape of Tumor Mutation Load, Mismatch Repair Deficiency, and PD-L1 Expression in a Large Patient Cohort of Gastrointestinal Cancers. Molecular cancer research : MCR 2018, 16(5):805-812.

20. Le Flahec G, Uguen M, Uguen A: Detection of Mismatch Repair Deficiency in Colorectal Cancers: Is It Really Time to Eliminate Immunohistochemistry? Journal of clinical oncology : official journal of the American Society of Clinical Oncology 2017, 35(3):376-377.

21. Uguen A, Gueguen P, Legoupil D, Bouvier S, Costa S, Duigou S, Lemasson G, Lede F, Sassolas B, Talagas M et al: Dual NRASQ61R and BRAFV600E mutation-specific immunohistochemistry completes molecular screening in melanoma samples in a routine practice. Human pathology 2015, 46(11):1582-1591.

22. Guan Q, Yan H, Chen Y, Zheng B, Cai H, He J, Song K, Guo Y, Ao L, Liu H et al: Quantitative or qualitative transcriptional diagnostic signatures? A case study for colorectal cancer. BMC genomics 2018, 19(1):99.

23. Guan Q, Chen R, Yan H, Cai H, Guo Y, Li M, Li X, Tong M, Ao L, Li H et al: Differential expression analysis for individual cancer samples based on robust within-sample relative gene expression orderings across multiple profiling platforms. Oncotarget 2016, 7(42):68909-68920.

24. Ao L, Zhang Z, Guan Q, Guo Y, Guo Y, Zhang J, Lv X, Huang H, Zhang H, Wang X et al: A qualitative signature for early diagnosis of hepatocellular carcinoma based on relative expression orderings. Liver international : official journal of the International Association for the Study of the Liver 2018, 38(10):1812-1819.

25. Chen R, Guan Q, Cheng J, He J, Liu H, Cai H, Hong G, Zhang J, Li N, Ao L et al: Robust transcriptional tumor signatures applicable to both formalin-fixed paraffin-embedded and fresh-frozen samples. Oncotarget 2017, 8(4):6652-6662.

26. Cheng J, Guo Y, Gao Q, Li H, Yan H, Li M, Cai H, Zheng W, Li X, Jiang W et al: Circumvent the uncertainty in the applications of transcriptional signatures to tumor tissues sampled from different tumor sites. Oncotarget 2017, 8(18):30265-30275.

27. Liu H, Li Y, He J, Guan Q, Chen R, Yan H, Zheng W, Song K, Cai H, Guo Y et al: Robust transcriptional signatures for low-input RNA samples based on relative expression orderings. BMC genomics 2017, 18(1):913.

28. Tejpar S, Stintzing S, Ciardiello F, Tabernero J, Van Cutsem E, Beier F, Esser R, Lenz HJ, Heinemann V: Prognostic and Predictive Relevance of Primary Tumor Location in Patients With RAS Wild-Type Metastatic Colorectal Cancer: Retrospective Analyses of the CRYSTAL and FIRE-3 Trials. JAMA oncology 2016.

29. Loupakis F, Yang D, Yau L, Feng S, Cremolini C, Zhang W, Maus MK, Antoniotti C, Langer C, Scherer SJ et al: Primary tumor location as a prognostic factor in metastatic colorectal cancer. Journal of the National Cancer Institute 2015, 107(3).

30. Salem ME, Weinberg BA, Xiu J, El-Deiry WS, Hwang JJ, Gatalica Z, Philip PA, Shields AF, Lenz HJ, Marshall JL: Comparative molecular analyses of left-sided colon, right-sided colon, and rectal cancers. Oncotarget 2017, 8(49):86356-86368.

31. Cha Y, Kim KJ, Han SW, Rhee YY, Bae JM, Wen X, Cho NY, Lee DW, Lee KH, Kim TY et al: Adverse prognostic impact of the CpG island methylator phenotype in metastatic colorectal cancer. British journal of cancer 2016, 115(2):164-171.

32. Mensenkamp AR, Vogelaar IP, van Zelst-Stams WA, Goossens M, Ouchene H, Hendriks-Cornelissen SJ, Kwint MP, Hoogerbrugge N, Nagtegaal ID, Ligtenberg MJ: Somatic mutations in MLH1 and MSH2 are a frequent cause of mismatch-repair deficiency in Lynch syndrome-like tumors. Gastroenterology 2014, 146(3):643-646 e648.

33. Cheng ML, Berger MF, Hyman DM, Solit DB: Clinical tumour sequencing for precision oncology: time for a universal strategy. Nature reviews Cancer 2018, 18(9):527-528.

34. Zhao W, Chen B, Guo X, Wang R, Chang Z, Dong Y, Song K, Wang W, Qi L, Gu Y et al: A rank-based transcriptional signature for predicting relapse risk of stage II colorectal cancer identified with proper data sources. Oncotarget 2016, 7(14):19060-19071.

35. Song K, Guo Y, Wang X, Cai H, Zheng W, Li N, Song X, Ao L, Guo Z, Zhao W: Transcriptional signatures for coupled predictions of stage II and III colorectal cancer metastasis and fluorouracil-based adjuvant chemotherapy benefit. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 2019, 33(1):151-162.

36. Qi L, Chen L, Li Y, Qin Y, Pan R, Zhao W, Gu Y, Wang H, Wang R, Chen X et al: Critical limitations of prognostic signatures based on risk scores summarized from gene expression levels: a case study for resected stage I non-small-cell lung cancer. Briefings in bioinformatics 2016, 17(2):233-242.

37. Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf U, Speed TP: Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 2003, 4(2):249-264.

38. Crans GG, Shuster JJ: How conservative is Fisher's exact test? A quantitative evaluation of the two-sample comparative binomial trial. Statistics in medicine 2008, 27(18):3598-3611.

39. Hochberg Y, Benjamini Y: More powerful procedures for multiple significance testing. Statistics in medicine 1990, 9(7):811-818.

40. Bland JM, Altman DG: The logrank test. Bmj 2004, 328(7447):1073.

Table 1. The Composition of 10-GPS

signature	gene1	gene2	signature	gene1	gene2
pair1	HNRNPL	CDC16	pair6	STRN3	TMEM192
pair2	MTA2	VGF	pair7	HPSE	BCAS3
pair3	CALR	SEC22B	pair8	PRPF39	ATF6
pair4	RASL11A	CAB39L	pair9	CCRN4L	GRM8
pair5	LYG1	DHRS12	pair10	AMFR	DUSP18

Notes:

A RCC sample was classified as MSI if the REOs (gene1>gene2) of at least 7 of the gene pairs in the 10-GPS vote for MSI; otherwise the MSS.

Table 2. The performances of the 10-GPS in RCCs of the independent datasets

	pre-MSI^a (MSI:MSS)^b	pre-MSS^a (MSI:MSS)^b	sensitivity	specificity	F-score
GSE39084_R	13(13:0)	18(0:18)	1	1	1
GSE18088_R	13(13:0)	15(1:14)	0.9286	1	0.9630
GSE75317_R	8(8:0)	18(1:17)	0.8889	1	0.9412
TCGA_R	15(9:6)	38(1:37)	0.900	0.8605	0.8798
Total_RCCs	49(43:6)	89(3:86)	0.9348	0.9348	0.9348

Notes:

^a represents the predicted MSI status by 10-GPS; ^b represents the original MSI status; GSE_R represents the RCC samples; Total_RCCs represents all the samples of RCC.

Table 3. The molecular characteristics of the five signature-disconfirmed RCC samples in the training dataset

original_MSI. status	predicted_MSI. status	KRAS. status	BRAF. status	CIMP. status
MSS	MSI	wild type	NA	NA
MSS	MSI	wild type	mutation	+
MSS	MSI	wild type	mutation	+
MSI	MSS	wild type	mutation	+
MSI	MSS	NA	wild type	-

Table 4: The datasets analyzed in this study from GEO and TCGA

	GSE39582 (n=566)	GSE39084 (n=70)	GSE18088 (n=53)	GSE75317 (n=59)	GSE13067 (n=74)	GSE13294 (n=155)	TCGA (n=457)
Stage
I	33	8	-	6	-	-	75
II	264	23	53	24	-	-	178
III	205	16	-	17	-	-	130
IV	60	22	-	12	-	-	64
Microsatellite status
MSI	75	16	19	11	11	78	11
MSS	444	54	34	48	63	77	81
Location
Right	224	31	28	26	-	-	261
Left	342	30	25	33	-	-	177
MSI_proportion
Right	57:154^a (27.0%)^b	13:18 (41.9%)	14:14 (50.0%)	9:17 (34.6%)	-	-	10:43 (18.9%)
Left	18:290^c (5.8%)^d	3:27 (10.0%)	5:20 (20.0%)	2:31 (6.1%)	-	-	1:33 (2.9%)
MSI detection	PCR	PCR or IHC	PCR	PCR	PCR	PCR	PCR
Adjuvant chemotherapy
Yes	233	-	-	-	-	-	-
No	316	-	53	-	-	-	-

Notes:

The data from GEO were produced by the same gene expression profiling platform (GPL570, Affy-HG-U133_Plus_2). ^arepresents the number of MSI and MSS of RCCs, respectively; ^b represents the proportion of MSI in RCCs; ^c represents the number of MSI and MSS of LCCs, respectively; ^d represents the proportion of MSI in LCCs.

Download PDF

Journal Publication

published 23 Oct, 2019

Read the published version in BMC Genomics →

Editorial decision: Major revision
16 Aug, 2019
Review #3 received at journal
08 Aug, 2019
Review #2 received at journal
06 Aug, 2019
Review #1 received at journal
06 Aug, 2019
Reviewer #3 agreed at journal
26 Jul, 2019
Reviewer #4 agreed at journal
26 Jul, 2019
Reviewer #2 agreed at journal
23 Jul, 2019
Reviewer #1 agreed at journal
01 Jul, 2019
Submission checks completed at journal
24 Jun, 2019
Editor invited by journal
24 Jun, 2019
Editor assigned by journal
24 Jun, 2019
Reviewers invited by journal
24 Jun, 2019
First submitted to journal
17 Jun, 2019

You are reading this older preprint version

Read the latest preprint version →

A Qualitative Transcriptional Signature for Predicting Microsatellite Instability Status of Right-sided Colon Cancer

Status:

Journal Publication

Version 1

Abstract

Figures

Background

Results

Discussion

Methods

Abbreviations

Declarations

Ethics approval and consent to participate

Consent for publication

Availability of data and materials

Competing interests

Funding

Authors' contributions

Acknowledgements

References

Tables

Supplementary Files

Status:

Journal Publication

Version 1